Identification of RIP
metallopeptidase RseP as Target Receptor for the Leaderless
Bacteriocin Ej97, and the
Presumptive Involvement of the Ecs ABC transporter in medium Resistance
Norwegian University of Life Sciences
Faculty of Veterinary Medicine and Biosciences Department of Chemistry, Biotechnology and Food Science
Master Thesis 2014 60 credits
Marianne Slang Jensen
I
Acknowledgements
The work implemented in this thesis was performed throughout 2014 at the Laboratory of Microbial Gene Technology (LMG) at the Norwegian University of Life Sciences (NMBU), with Professor Dzung Bao Diep and Professor Helge Holo as supervisors.
I would like to begin by thanking Dzung for accepting me as his Masters student, and for being ever enthusiastic and inspiring in his guidance. You have always met me with genuine interest, some bad jokes, and invigorating thoughts and discussions in abundance.
Several lab members deserve special thanks for helping me complete my thesis. This project is a continuation of the work of Kirill Ovchinnikov, who throughout this year has indulged me with interesting perspectives, suggestions, and a bit of criticism when appropriate. I would also like to thank Ibrahim Mehmeti who helped me get my project off to the best possible start. Özgün Candan Onarman Umu, thank you for all your patience with me in the preparations of whole genome sequencing. You are a perfectionist far surpassing me and I could not have wished for a better teacher. Cyril Frantzen, I am grateful for all your help with the MiSeq, I could not have done it without you and would quite possibly still be stuck at this point. Thank you for devoting your time to my silly beginner’s questions and guiding me through the analysis process.
Thanks to all the people at LMG for a great year which I have sincerely appreciated more than I could have imagined. Little could go wrong with Linda Godager ensuring the lab was always in order. May-Britt Selvåg Hovet, thank you for tirelessly wielding the autoclave, but also for making me feel right at home in the lab. Hai Chi, Eirik, Juan José, Andreza, Anne Kristin and Pawel, thank you all for good advice, for social events, and for a great working environment. I will be sad to leave you.
Finally I would like to thank my family and friends for all their support and encouragement.
Ås, December 2014
Marianne Slang Jensen
II
Abstract
In a recent study on the leaderless and Lactococcus-specific bacteriocin LsbB, the results implied that LsbB utilizes a zinc-dependent metallopeptidase YvjB from the M50 superfamily as receptor for target recognition and subsequent killing of other lactococcal bacteria.
Based on this evidence, another study was initiated which elucidated a conserved motif near the C-terminal of LsbB, proving to be crucial for its activity. The work also revealed that several other leaderless bacteriocins, including Ej97, K1 and EntQ, all share this conserved motif with LsbB, suggesting that they target the same receptor. Despite this similarity, these bacteriocins display a significantly broader inhibition spectrum than LsbB.
For this thesis, it was initially affirmed that Ej97 and K1 exhibit a wide inhibition spectrum relative to LsbB, followed by generation of approximately 80 spontaneous mutants resistant to Ej97, K1 or EntQ. These mutants belonged to pediococcal, enterococcal and lactococcal strains.
The homologous genes of the LsbB receptor (rseP) were identified in order to be sequenced and analyzed in all the collected mutants. Severe mutations inflicting premature termination of RseP were in many cases revealed. This suggests that these RseP proteins which are closely related to the LsbB receptor also serve as docking molecules in other bacterial species for LsbB-like bacteriocins. They are however unable to function as receptor for LsbB.
Still, a large proportion of the mutants did not hold mutations within the putative RseP receptor.
Phenotypic microtiter assays were performed to investigate Ej97 resistance levels. It became evident that the degree of resistance generally diverged into two levels – one very high indicating absolute resistance, and a lower level but still significantly higher than in the wild type.
Interestingly, the mutants with the highest resistance level coincided with the presence of dysfunctional RseP.
Concluding the work for this thesis, whole genome sequencing of 40 mutants was carried out to identify possible reasons for the resistance observed in mutants with intact RseP and lower level of resistance. Variant detection provided clear indications of components from the Ecs ABC- transporter being involved. After this revelation it was noticed that previous studies on non- bacteriocin systems have indicated a connection between EcsAB and RseP in other species.
Future work will examine how these genes are interrelated in conferring bacteriocin resistance.
III
Sammendrag
I et nylig studie på det lederløse og laktokokk-spesifikke bakteriosinet LsbB ble det konstantert at LsbB anvender en sink-avhengig metallopeptidase YvjB fra M50-superfamilien som reseptor for å gjenkjenne og angripe målceller.
Et påfølgende studie viste at en rekke andre lederløse bakteriosiner, inkludert Ej97, K1, og EntQ, deler et konservert motiv ved C-terminalen som er essensielt for aktiviteten hos LsbB. Dette tyder på at de benytter seg av den samme reseptoren, men til tross for denne likheten har disse slektningene betydelig bredere inhibisjonsspektra sammenlignet med LsbB.
I denne avhandlingen ble det i første omgang bekreftet at Ej97 og K1 har et bredt inhibisjonsspektrum i forhold til LsbB, og det ble deretter generert omkring 80 spontane
mutanter resistente mot Ej97, K1 eller EntQ. Disse mutantene tilhørte pediokokke, enterokokke og laktokokke stammer. I hver av de aktuelle stammene ble det homologe genet til LsbB- reseptoren (rseP) identifisert. rseP kunne deretter bli amplifisert og sekvensert i samtlige mutanter, og det ble i mange tilfeller oppdaget alvorlige mutasjoner som medførte pre-
terminering av RseP. Dette tydet på at disse RseP-membranproteinene som er nært beslektet med LsbB-reseptoren også kan fungere som reseptor i andre bakteriearter for LsbB-lignende
bakteriosiner. Samtidig kan de derimot ikke fungere som reseptor for LsbB.
Imidlertid var det en stor andel av mutantene som ikke hadde mutasjoner innad det antatte reseptorgenet. Fenotypiske microtiter assays ble utført for å se på resistensgraden mot Ej97. Det viste seg at graden av resistens generelt kunne deles i to nivåer – ett svært høyt som indikerte full resistens, og et lavere men samtidig klart høyere resistensnivå enn hos villtypen. Mutantene med høyest resistens var de samme som hadde mutasjoner i rseP.
Avslutningsvis ble det foretatt helgenomsekvensering av 40 mutanter for å lokalisere mulige årsaker til resistensen man observerte i mutanter med inntakt RseP og lavere resistensnivå.
Variantdeteksjon ga klare indikasjoner på at komponentene av Ecs ABC-transportøren er involvert. Etter denne oppdagelsen ble det bemerket at tidligere studier ikke relatert til
bakteriosiner har avdekket sammenhenger mellom EcsAB og RseP i andre bakteriearter. Videre
arbeid vil se nærmere på hvordan disse genene henger sammen i forhold til bakteriosinresistens.
Table of Contents
1. INTRODUCTION... 1
1.1 Antimicrobial agents ... 1
1.2 Bacteriocins ... 3
1.2.1 Applications and future prospects ... 4
1.2.2 Biosynthesis and mode of action ... 6
1.2.3 Resistance ... 10
1.3 Classification of bacteriocins ... 11
1.4 Bacteriocins addressed in this study ... 15
1.4.1 Leaderless bacteriocin LsbB... 15
1.4.2 Leaderless enterocins Ej97, K1 and EntQ ... 17
1.4.3 Sequence similarity ... 17
1.5 Bacteriocin receptors ... 18
1.6 RseP of the M50 superfamily ... 23
1.7 The aim of this study ... 24
2. MATERIALS ... 25
2.1 Growth media and agars ... 25
2.2 Bacterial strains ... 25
2.3 Synthetic peptides ... 27
2.4 Laboratory equipment ... 27
2.5 Instruments ... 28
2.6 Software ... 29
2.7 Kits ... 29
2.8 Chemicals and reagents ... 31
2.9 Enzymes ... 32
2.10 DNA and standards ... 32
2.11 Primers ... 32
3. METHODS ... 34
3.1 General methods in microbiology ... 34
3.1.1 Sterile working technique ... 34
3.1.2 Preparation of growth medium and agars ... 34
3.1.3 Streaking bacteria onto agar plates ... 34
3.1.4 Inoculation and cultivation of overnight pure cultures ... 34
3.1.5 Long-term storage of bacteria ... 35
3.2 Schematic of work progression ... 35
3.3 Spot-on-lawn inhibition spectrum assays ... 36
3.4 Accumulation of bacteriocin resistant mutants ... 38
3.5 Microtiter plate phenotype assays ... 38
3.6 DNA isolation for genotype assays ... 40
3.6.1 DNA isolation for single-gene analysis ... 40
3.6.2 DNA isolation for whole genome sequencing... 42
3.7 DNA measurements ... 43
3.7.1 NanoDrop ... 43
3.7.2 Qubit ... 43
3.8 Primer design and preparation ... 44
3.9 Polymerase chain reaction (PCR) ... 45
3.10 Agarose gel electrophoresis... 47
3.11 PCR product clean-up ... 49
3.12 Sequencing ... 49
3.12.1 rseP single gene sequencing ... 50
3.12.2 Whole genome sequencing on Illumina MiSeq ... 50
3.13 Analysis of sequence data ...58
3.14 Unspecific resistance assay ... 65
3.15 Mutation frequency with mutation specific primers (E. faecalis 3358) ... 65
4. RESULTS ... 67
4.1 Inhibition spectrum assays ... 67
4.2 Collection of bacteriocin resistant mutants ... 69
4.3 Microtiter plate assays ... 71
4.4 rseP analysis ... 73
4.5 Whole genome sequence analysis ... 78
4.6 Assessment of resistance specificity ... 83
4.7 Mutation frequency (E. faecalis 3358) ... 83
5. DISCUSSION ... 84
5.1 Inhibition spectra ... 84
5.2 The RseP receptor ... 85
5.3 Phenotype of naturally resistant mutants ... 88
5.4 Resistant RseP genotypes ... 89
5.5 Whole genome sequencing ... 90
5.6 Involvement of the Ecs ABC transporter in resistance ... 91
5.7 Mutations identified give rise to specific resistance ... 93
5.8 A summary of overall results ... 94
5.9 Concluding remarks and future prospects ... 96
6. REFERENCES ... 98
APPENDIX ... i
1
1. INTRODUCTION
1.1 Antimicrobial agents
Physical and chemical preservatives such as decreased pH, high salt concentration, low water content, and intense temperatures have been employed in foods for suppressing microbial growth for thousands of years. Stronger chemicals and treatments have additionally been exploited for decades in disinfection and sterilization to prevent crossover of contamination to foods and humans, but are generally not compatible with human consumption. These substances include components such as phenol, ethanol, chlorine, and peroxide, as well as newer compounds with prolonged activity. Among the utilized physical treatments are various heat- and radiation-based processes (APUA 2014; Todar 2008).
The first chemotherapeutic agents for fighting infectious diseases were certain alkaloids, i.e.
plant extracts with strong physiological effects, and were put to use in the 1600s in South
America primarily to treat malaria. Synthetic agents such as the arsenic-containing salvarsan was in the early 1900s developed to cure syphilis, but due to the toxicity it was not ideal for
treatment. In the 1930s, the introduction of sulfonamides drastically improved treatment of infectious diseases, many of which are still being used today (Yazdankhah et al. 2013).
Bacterially produced substances with antagonistic activity comprise a large group, including traditional antibiotics, bacteriophages and antimicrobial peptides (AMPs), as well as various other inhibitory agents such as metabolic by-products (ammonia, hydrogen peroxide etc.) and bacteriolytic/autolytic enzymes (Jack et al. 1995; Tagg et al. 1976).
Traditional antibiotics have been utilized since the 1940s for targeting pathogens causing infectious diseases in humans and animals. Being discovered by Alexander Fleming in 1928 by accidental mold contamination on staphylococci, Howard Florey and Ernst Chain were
responsible for putting the first antibiotic (penicillin) into large-scale production by 1942. The
three scientists were later awarded the Nobel Prize in Physiology for their achievements. The
new “miracle cure” was soon distributed to the public, proving especially valuable to soldiers
during World War II in the following years. However, the limitations of penicillin were soon
2
uncovered, with problems concerning both the activity range as well as causing allergic reactions in some patients. Large sums of money were subsequently invested in discovering new types of antibiotics, initiating the antibiotic golden era. However, in the 1960s the development came to a sudden halt (Yazdankhah et al. 2013).
Since then, only one new antibiotic class (the oxazolidinones) have been discovered, and the lack of development within the field of traditional antibiotics for the past decades is not an ideal situation for maintaining effective treatment options (Hassan et al. 2012). An increasing problem concerning traditional antibiotics is the emergence of resistant and multi-drug-resistant (MDR) bacteria in response to prolonged and inadequate use. Poor practice is widespread in human healthcare due to overprescribing, but moreover in animal feeds for controlling disease and promote increased growth in large populations. This poses a threat of resistance crossing from animal products to humans. Supplementation of antibiotics in feed has thus been banned in several countries to restrict dispersion of resistance and MDR development, but the prevalence of resistant bacteria have already reached a level of great concern (Cheng et al. 2014).
Resistance develops and spreads by genetic mutations rendering a bacterium unsusceptible to the actions of the antibiotic, and the resistance-determinants subsequently being transferred between bacteria. This process is assisted by horizontal gene transfer through conjugation (transfer of genetic material between bacteria through cell-to-cell contact), transduction (facilitated by host DNA-uptake in virus) and transformation (incorporation of free DNA from the environment by competent cells). Together, these mechanisms contribute to restricting treatment options in patients and have severe consequences in regards to mortality rates and economic perspectives, e.g. the increasing occurrence of methicillin-resistant Staphylococcus aureus (MRSA) infections in hospitals and nursing homes (Huddleston 2014).
Many alternatives to traditional antibiotics have been proposed and are presently being
investigated for various applications in animal feed, food preservation, and the pharmaceutical
industry. They include antibacterial vaccines, immunomodulatory agents, bacteriophages and
their lysins, probiotics, plant extracts, inhibitors of bacterial quorum sensing/biofilm/virulence,
feed enzymes, and antimicrobial peptides (AMPs) (Cheng et al. 2014).
3
1.2 Bacteriocins
For millennia, human populations across the world have benefitted from the fermentative
metabolism of lactic acid bacteria (LAB) and the resulting pH decrease, in order to prevent rapid microbial spoilage of foods. In addition to this general inhibition mechanism, scientists of the past century have revealed that most LAB also possess additional properties constituting mechanisms of a significantly more specific nature. These mechanisms involve the production and secretion of small antimicrobial peptides, termed bacteriocins, presumed to attach to unique receptor molecules on the target in order to exert potent activity. The bacteriocins are not to be confused with the colicins and microcins, which although also proteinaceous substances are produced and mainly active against Gram-negative species. These substances and their possible applications will not be discussed in detail here (Eijsink et al. 2002; Nes et al. 1996).
Production of AMPs is a widespread trait within both the prokaryotic and eukaryotic domain, but while contributing to the innate immunity system of eukaryotic life forms for protection against invading organisms, AMPs produced by bacteria are primarily considered as a means of strategic warfare between competing bacteria within the same niche (Nes et al. 2007; Nissen-Meyer &
Nes 1997). The abundant use of AMPs in defense mechanisms has likely been evolutionally successful due to ribosomal synthesis of peptides being a rapid response process compared to e.g. production of antibiotics, antibodies, and immune cells (Nissen-Meyer & Nes 1997).
In the broadest sense of the word, bacteriocins are defined as ribosomally produced peptides with antimicrobial activity. The bacteriocins should be regarded as antimicrobial compounds separate from the traditional antibiotics due to being gene-encoded, while traditional antibiotics are secondary metabolites produced by the assistance of multi-enzyme complexes (Nes et al. 2007).
Additionally, in contrast to traditional antibiotics which usually act as enzyme inhibitors (causing damage to cell-wall synthesis, DNA synthesis, protein synthesis, or other metabolic pathways), bacteriocins are generally known to exert their activity on target cells by permeabilization of the cell membrane causing leakage and loss of ion gradient (Diep et al. 2007).
By observing the inhibitory effects caused by bacteria isolated from urine (presumably E. coli)
on the pathogen Bacillus anthraces (causing anthrax), Pasteur and Joubert were likely the first to
record such antagonistic interactions between bacteria (Tagg et al. 1976). Recognition of these
4
antimicrobial agents as secreted peptides, later termed colicins, was first demonstrated by Gratia in 1925 when he observed that the E. coli strain V produced a heat-stable substance that
effectively inhibited the growth of another E. coli strain ϕ (Reeves 1965; Tagg et al. 1976).
Discovery of several other colicins and improved knowledge of their characteristics ensued in the following years, and the more general term “bacteriocin” was appointed by Jacob et al in 1953 after it had been established that these substances could be produced by other bacteria than coliform members of the Gram-negative Enterobacteriaceae family (Jack et al. 1995).
The characterization of bacteriocins from Gram-positive LAB species was heavily based on prior knowledge of the colicins. This included being proteinaceous substances with a relatively narrow inhibition spectrum and intraspecies killing activity at low concentrations, and which depended on the presence of a specific receptor on the target. Additionally, they were known to be resistant to heat inactivation and produced along with substances which rendered the producer itself immune. Their proteinaceous character could be affirmed by observing inactivation of the substances by certain proteases, as well as trypsin treatment being capable of rescuing susceptible target cells after having been exposed to a lethal dose of bacteriocin. The latter experiment also suggested a two-step mechanism, with the bacteriolytic activity being initiated from the receptor site after the bacteriocin had already docked. This conclusion was assumed after observing that trypsin rescue was possible only if initiated within the first couple of minutes after bacteriocin exposure (Reeves 1965; Tagg et al. 1976).
New knowledge has revealed that the bacteriocins of the Gram-positive bacteria comprise a highly heterogeneous group with a generally lower molecular weight than the colicins and displaying a somewhat less narrow activity spectrum, in some cases being active even towards Gram-negative species (Jack et al. 1995; Tagg et al. 1976).
1.2.1 Applications and future prospects in food safety and pharmaceutical industries
Nisin is presently the sole example of a commercially utilized bacteriocin widely implemented in
food preservation. The peptide has been known since 1928 as a product of Lactococcus lactis
strains, and was structurally characterized already in 1971 (Jack et al. 1995). By then, nisin-
producing strains had already been implemented in dairy production for a long time without
occurrence of any adverse effects. In 1988, nisin was approved for use as a food additive in the
5 US food industry after the Food and Drug Administration allowed nisin produced by certain strains to be employed to prevent outgrowth of Clostridium botulinum spores and toxin formation in pasteurized cheese spreads and canned foods (FDA, 1988).
The target cells of bacteriocins are generally confined to a relatively narrow selection of closely related LAB, but frequently also include prominent Gram-positive food spoilage bacteria and human pathogens. The peptides (although pH-tolerant and heat-stable) are easily degraded in the body and have no effect on human cells due to their distinct differences in composition relative to bacterial cells. Like their producers, they are thus generally recognized as safe substances (GRAS-status) for human consumption and handling. Additionally, many of the LAB known to produce bacteriocins which inhibit the growth of food spoilage bacteria are often found naturally in our intestine or already widely implemented in the food industry. These are all exceedingly attractive traits, establishing the LAB bacteriocins as promising tools for future industrial
exploitations and safe applications in food preservatives and pharmaceutics (Eijsink et al. 2002).
Additionally, as the bacteriocins are ribosomally synthesized from dedicated genes and often contain minimal post-translational modifications, they are interesting subjects in peptide
engineering for enhancing traits. Engineering inducing various modifications, gene-shuffling and re-design can possibly enhance the activity, potency and stability of the peptide. For this to be successfully executed with rational design, it is necessary to have profound knowledge of the structure-function relationship. However, modern techniques allow scientists to do high- throughput screening of random mutants and hybrids, facilitating directed evolution on a large scale (Jack et al. 1995; Lohans & Vederas 2012; Nes et al. 1996).
The administration of possible commercial bacteriocins could be executed in a variety of ways, including as purified additives or produced in situ (e.g. in bacterial starter cultures or as
probiotics). In many cases it could prove favorable to administer the peptides in combination with other preservatives and antimicrobials, or in combination with other bacteriocins. This combination-strategy could potentially enable more natural preservation of foods with less need of chemicals and physical treatments (Galvez et al. 2007). LAB bacteriocins are commonly not active against Gram-negative species due to the outer membrane forming a protective barrier.
However, by using a combination of bacteriocin and a chelating agent such as EDTA in
simultaneous treatment for weakening the outer membrane, there is a possibility of expanding
6
the applications of bacteriocins and increase their value in the food industry. Trials with nisin have provided encouraging results in regards to such opportunities (Stevens et al. 1991).
Regardless, bacteriocins need to be vigorously tested in clinically relevant settings to evaluate their real potential in pharmaceutics, farming, and food industry. Additionally, the efficacy of bacteriocin synthesis is influenced by several environmental factors, and therefore the optimal conditions must also be determined for individual bacteriocins to make them economically viable in industries. For commercial use to be feasible, methods for large-scale production and
purification with high yield must be further developed and improved (Galvez et al. 2007).
Another valuable trait of the bacteriocins are their activity spectra which range from very narrow to fairly wide, and it has long been known that strains frequently produce more than one
bacteriocin. In treatment of infectious diseases, it is often preferable to employ narrow-spectrum agents in order to prevent disrupting the balance of the commensal gut-community, a problem often associated with traditional broad-spectrum antibiotics and which assist the emergence of opportunistic pathogens, capable of taking advantage of the sudden lack of competition.
However, when the illness is of unknown origin, the better choice is rather broad-spectrum agents, and many known bacteriocins hold this trait as well. Due to bacteriocins targeting unique sites on susceptible cells and conducting their mode of action as permeabilization of the
membrane, the probability of cross-resistance to traditional antibiotics is generally low. Several bacteriocins have already displayed promising activity against antibiotic resistant strains such as vancomycin-resistant enterococci (VRE) and opportunistic Staphylococcus aureus (MRSA) (Hassan et al. 2012; Lohans & Vederas 2012).
Like traditional antibiotics, it is likely that problems with resistance will occur with bacteriocins if they are put into commercial use. However, scientific developments since the introduction of traditional antibiotics can assist todays scientists in preforming the necessary studies for fully understanding the resistance mechanisms and thereby minimize this problem (Cotter et al. 2013).
1.2.2 Biosynthesis and mode of action
Bacteriocin production is an energy consuming and thus often strictly regulated process. The peptides are generally produced from dedicated loci with leader peptides attached up until
processing and secretion from the cell. It is generally believed that bacteriocins kill other bacteria
7 by pore-formation in the membrane after receptor recognition, while the producers themselves remain immune to the bacteriocin activity by expressing specific immunity proteins.
Bacteriocin loci: In general, loci dedicated to bacteriocin production are dependent on several elements in addition to the structural bacteriocin genes; including components for processing and secretion, regulation, and self-immunity. The structural bacteriocin genes are usually located upstream the corresponding immunity genes and co-regulated with these, as illustrated in figure 1.1 (Eijsink et al. 2002). Commonly, a minimum of four genes are essential for bacteriocin production and are usually found in close proximity within the same loci. The structural genes encode pre-bacteriocins which have yet to be cleaved and folded, the immunity genes provide protection for the producer, dedicated membrane-bound ABC transporters facilitate exportation, and an accessory protein with elusive function is also needed for secretion. Exceptions to this organization are the class IIb bacteriocins whose activity require the action of two different peptides (two structural genes), and the class I bacteriocins (lantibiotics) which need additional genes for catalyzing dehydration of selected amino acids and lanthionine ring formation, and serine proteases to cleave off the leader peptide (Nissen-Meyer & Nes 1997).
Figure 1.1: Organization of the class IIa bacteriocin sakacin A sap locus. sapA encodes the bacteriocin precursor, and saiA the cognate immunity protein. orf4 produces a Sap-Ph pheromone precursor. sapK codes for a sensory histidine protein kinase, sapR for the response regulator, sapT for the ABC transporter, and
sapE for the cognate accessory protein. IS is an insertion sequence element, and the triangles marked “i” and“ii” indicate regulated promotors (Diep et al. 2000).
Production: Genetic determinants for bacteriocin production are occasionally chromosomally
located, but more commonly plasmid-bound or even transposon-associated. As mobile genetic
elements they are capable of promoting transfer to other bacteria by horizontal gene transfer. The
status of bacteriocin synthesis rely on the physiological conditions surrounding the producer,
including growth media composition, temperature, incubation time, pH, and aeration. These
factors must be empirically investigated to establish optimal conditions for triggering bacteriocin
synthesis. Production can in some cases also be induced by the assistance of certain substances
or irradiation with ultraviolet light (Jack et al. 1995; Tagg et al. 1976).
8
Processing and secretion: Most bacteriocins are produced as linear pre-peptides with a leader sequence extension at the N-terminal, commonly in the form of a 14-30 aa double-glycine leader.
This extension is responsible for maintaining inactivity of the peptide whilst inside the producer cell. The leader peptide is cleaved off upon bacteriocin secretion by dedicated ABC-transporter machinery, or in some cases by a sec-dependent pathway. The accessory protein is also essential for successful externalization of the peptide, although its specific role has yet to be established.
After cleavage and modifications, the peptide reaches mature and active state after having been transported to the extracellular environment and attained its three dimensional folded structure (Eijsink et al. 2002; Nes et al. 1996).
Self-immunity: Genes conferring self-immunity are crucial to protect the producer from being harmed by the bacteriocin. They are commonly co-regulated with the bacteriocin and hold a conserved position downstream of the structural bacteriocin genes, making them relatively easy to identify. Still, most mechanisms of self-immunity remain elusive (Kjos et al, 2009). Studies indicate that immunity is conferred in various manners, with the presently best studied models being nisin and class IIa bacteriocins. Immunity in nisin-producing strains is conferred by two separate mechanisms, namely a specialized ABC transporter which pumps nisin out of the membrane, and a dedicated immunity protein (LanI) which interacts with nisin on the
extracellular side and prevents it from binding to the receptor (Draper et al. 2008). In the case of class IIa producers, the results suggest that the immunity proteins bind tightly to the intercellular part of the receptor and thereby prevents pore formation in the presence of bacteriocin (Diep et al. 2007; Hassan et al. 2012).
Regulation: As processing and secretion of bacteriocins is dependent on ABC transporters and their ATP-binding component for coupling ATP hydrolysis with transportation, the production is energy-demanding and often strictly regulated to avoid wasting resources. Some bacteriocins, especially those of Gram-positive species, are known to be regulated according to cell-density in a phenomena known as quorum-sensing. By using secretion of pheromones to communicate with members of the same strain, the bacteria can co-ordinate production of bacteriocin (Nes et al.
1996). This mechanism enables the bacterium to sense its own growth compared to competing bacteria, ensuring that full bacteriocin production is only initiated when it is useful to the cell, i.e.
when the environment contains a certain density (quorum) of neighboring cells. Class I
9 bacteriocins utilize the bacteriocin itself as inducing factor and is thus auto-regulated by the bacteriocin. Regulation of class II bacteriocins by quorum-sensing is dictated by pheromone induction instead, with the pheromone being structurally similar to the bacteriocin but without the antimicrobial activity (Snyder & Worobo 2014). This quorum-sensing mechanism is frequently termed three-component regulatory system. The system consists of three co- transcribed genes which encode a bacteriocin-like peptide (pheromone), a specific histidine kinase (sensor protein), and a DNA-binding effector protein (response regulator). Activation of bacteriocin production commences with secretion of the bacteriocin-like pheromone, which binds to the histidine kinase sensor protein in the membrane. Bacteria of the same strain will all secrete this pheromone at low levels, and when being in sufficient numbers reaching a certain threshold level, a chain reaction of auto-phosphorylation will be initiated. The phosphoryl group will during this process be transferred to the response regulator, which in turn becomes capable of binding DNA and activate the regulated promotor of the bacteriocin locus (figure 1.2). This induces massive pheromone and bacteriocin production (Eijsink et al. 2002).
Figure 1.2: Schematic of how bacteriocin production can be regulated by quorum-sensing. The inducing
factor IF (bacteriocin-like pheromone) binds to the N-terminal sensory domain of histidine kinase protein HK
at the exterior of the membrane. This stimulates activation of the C-terminal autokinase activity of HK on the
inside of the membrane. This activation initiates auto-phosphorylation, leading to the response regulator RR
being phosphorylated and inducing transcription of bacteriocin-affiliated operons (Nes et al. 1996).
10
Initial interaction: Contact between extracellular bacteriocin and target cell membrane is
facilitated by unspecific electrostatic interactions due to the positively charged cationic character of the bacteriocin and the negatively charged anionic lipids of the target membrane (Eijsink et al.
2002; Nissen-Meyer & Nes 1997). Evidence of the bacteriocins’ usually amphiphilic structure suggested that the killing activity was enabled by the hydrophobic regions interacting with the target cell membrane and causing pore-formation (Nissen-Meyer & Nes 1997; Tagg et al. 1976).
Target recognition and mode of action: The specific interaction occurs when the bacteriocin recognizes and docks with a specific receptor on the surface of the target cell. The spectrum of activity is thus determined by which bacteria contain the receptor molecule compatible with the bacteriocin. The common mode of action is so far known to be attachment to the receptor, and subsequently initiating pore-formation in the membrane. This leads to leakage and dissipation of the proton motive force, causing lysis of the target bacterium. Although this is the general mechanism, the resulting pores can differ greatly in respect to size, stability and conductivity.
Additionally, the receptor-binding and pore-formation is affected by the membrane potential and pH, implying that physiological state of the target cell also influences sensitivity. Alternative inhibition mechanisms to pore-formation have also been identified, e.g. by binding and removing lipid II from its functional site, hence inhibiting cell wall synthesis (Eijsink et al. 2002).
1.2.3 Resistance
With the early knowledge of the colicins and the notion that they attach and interact with specific receptor molecules on the surface of target cells, it was proposed that the observed resistance to colicins was due to loss of this receptor, i.e. by genetic mutations rendering the receptor unable to absorb the peptide. These receptor mutations often result in loss or poor affinity between receptor and bacteriocin. One should note that resistance is distinct from immunity, where the latter involves specific proteins protecting the producer from the activity of the bacteriocin (Hassan et al. 2012; Reeves 1965; Tagg et al. 1976). Resistance may be caused by several events, including 1) mutations altering the receptor itself, 2) mutations elsewhere affecting the
placement, orientation or availability of the receptor binding site, 3) mutations causing unspecific resistance by increased afflux pump exportation or protease activity degrading bacteriocins, or 4) mutations conferring a decrease in receptor expression and thus fewer available receptor
molecules (Cotter et al. 2013; Hassan et al. 2012; Lohans & Vederas 2012).
11 As proposed in event 2, altering the composition and structure of the cellular membrane may render a bacterium resistant to a bacteriocin by restricting the peptide from reaching its target.
The physiological state of the cell can as previously mentioned also affects the effectiveness pore-formation (Eijsink et al. 2002). Additionally, lack of bacteriocin sensitivity may be a result of non-specific immunity proteins due to these proteins not always being completely specific, and thus can provide protection towards several bacteriocins. This property of cross-immunity is difficult to explain as the sequence of the immunity genes display considerable variation (Cotter et al. 2013; Eijsink et al. 2002).
1.3 Classification of bacteriocins
Bacteriocins commonly consist of 20-70 amino acids (aa) and display large variation in terms of producer species, composition, physical properties, secretion mechanism, post-translational modifications (PTMs), receptor, and inhibition spectra. Bacteria are additionally often capable of producing multiple types of bacteriocins. This complicates classification, and the issue is still under debate. New research has and will continue to alter the concept of bacteriocin
classification to best represent present knowledge (Nes et al. 1996; Snyder & Worobo 2014).
Table 1.1: Classification scheme for the bacteriocins of Gram-positive bacteria. There are four main classes of bacteriocins which are further divided into various subgroups. Antimicrobial peptides produced by Gram- negative bacteria make up an entirely separate group consisting of microcins and colicins (Nes et al. 2007).
Producer bacteria Peptide term Class Subclass Description Example
Gram-negative
producers
Microcins Colicins
<10 kDa
>30 kDa Gram-positive
producers
Bacteriocins Class I Type A Type B Type C
Elongated and flexible Globular and rigid Multicomponent
Nisin Mersacidin Staphylococcin C55 Class II Subclass IIa
Subclass IIb Subclass IIc Subclass IId Subclass IIe
Pediocin-like motif Two-peptide
Linear, non-pediocin like Leaderless
Protein-derived
Pediocin PA-1 Lactococcin G Lactococcin A LsbB
Propionicin F Class III IIIa
IIIb
Bacteriolytic Non-bacteriolytic
>30 kDa heat-labile proteins
Class IV Cyclic bacteriocins Garvicin-ML
12
Microcins and the larger colicins will not be addressed in detail in this thesis. These are ribosomally synthesized peptides like the bacteriocins, but are produced by Gram-negative species. They generally have a more narrow inhibition spectrum compared to the bacteriocins of the Gram-positive bacteria (Duquesne et al. 2007; Nissen-Meyer & Nes 1997). Bacteriocins produced by Gram-positive bacteria are normally not active against Gram-negative bacteria, and vice versa. The protein-sized colicins secreted by E. coli were for many years best studied, but the peptide-sized bacteriocins from LAB have become increasingly more interesting in recent years as good candidates for applications in various industries (Nes et al. 2007).
Class I contains the lantibiotics which are defined by their amino acids being subject to distinctive post-translational modifications (PTMs) after the peptides are ribosomally
synthesized. The peptides undergo extensive PTMs such as dehydration of selected serine and threonine residues resulting in altered amino acids including 2,3-di dehydroalanine, D-alanine, 2,3-didehydrobutyrine, and formation of the characteristic lanthionine ring-structures (Eijsink et al. 2002; Hechard & Sahl 2002). Nisin was the first lantibiotic to be characterized and today more than 50 other lantibiotics have been identified. Several of these show promising activity towards MRSA and VRE. The lantibiotics are commonly divided into type A and type B
lantibiotics, as well as certain two-peptide lantibiotics. Nisin is categorized as type A with its 34 amino acids and five lanthionine rings (figure 1.3). Type B lantibiotics are generally shorter and not able to facilitate pore-formation as they cannot stretch across the membrane. Despite this, they also utilize lipid II for target recognition (Martin & Breukink 2007). The lantibiotics of subgroup A are elongated, amphiphilic and positively charged, while subgroup B are globular and non-charged (Nes et al. 2007; Nissen-Meyer & Nes 1997).
Figure 1.3: Illustration of the sequence and structure of nisin, displaying the placement of its five
characteristic lanthionine rings A-E (Martin & Breukink 2007).
13 Class II consists of primarily non-modified peptides, and due to this they have simpler structures lacking PTMs (except for disulfide bridges or circularization, see below) as well as the genes responsible for performing distinct PTMs. Class II bacteriocins are produced by many LAB and comprise a large group, frequently divided further into five subgroups (a-e) (Nes et al. 2007).
Class IIa bacteriocins are recognizable by a pediocin-like consensus motif YGNGV and a disulfide bridge in their highly conserved hydrophilic and cationic N-terminal. The N-terminal folds into a secondary structure of antiparallel β-sheets, separated from the C-terminal by a hinge structure. The C-terminal has a secondary structure of an α-helix, and consists of a less
conserved hydrophobic/amphiphilic domain involved in target specificity and membrane permeabilization (Nes et al. 2007; Nissen-Meyer & Nes 1997). The class IIa peptides consist of between 37 and 48 amino acid residues, and are known for their strong antilisterial effect.
However, their activity spectra also include strains from various Lactobacillus, Pediococcus, Enterococcus, Carnobacterium, Leuconostoc, Lactococcus and Clostridium species (Drider et al.
2006; Eijsink et al. 2002; Fimland et al. 2005).
Figure 1.4: The folded structure of class IIa bacteriocin Leucocin A in solution (determined by NMR), displaying its N-terminal β-sheets and C-terminal α-helix secondary structure (Lohans & Vederas 2012).
Class IIb contains the two-peptide bacteriocins which depend on the combined action of two
peptides to reach full antimicrobial activity. In some cases the individual peptides are completely
inactive, while other peptides show some activity individually but combined reveals a strong
synergetic effect. One-peptide bacteriocins might as well display synergetic activity when
combined, but unique to the IIb bacteriocins is that their genes are located in the same operon
(Eijsink et al. 2002). Additionally, the producers only express one single immunity protein,
which is also located within the same operon. It has been established that the peptides attain
close interaction when they inflict antibacterial activity, but the configuration of the peptides as
14
heterodimers is still elusive. Class IIb bacteriocins exhibit quite narrow activity spectra, and with each peptide being relatively short at 25-38 aa. Two-peptide bacteriocins also exist within the lantibiotics, but are as previously elaborated subjected to specific PTMs (Garneau et al. 2002).
Class IIc represents diverse and unsorted bacteriocins that do not fit into the other subgroups.
They are non-lantibiotic, non-pediocin-like, and one-peptide bacteriocins (Eijsink et al. 2002).
Class IId are leaderless peptides, i.e. produced without leader sequence. Most bacteriocins are attached to a N-terminal leader peptide after translation, which needs to be cleaved off upon leaving the producer cell for the bacteriocin to reach functional state. The leader sequence- extension at the N-terminus is crucial to maintain inactive state whilst inside the producer, and how leaderless bacteriocin producers protect themselves from the antimicrobial activity is still unknown (Nes et al. 2007).
This study is based upon members of this subclass, and will be elaborated in more detail during subsequent sections.
Clas IIe bacteriocins are protein-derived bacteriocins produced from larger proteins by specific degradation mechanisms (Nes et al. 2007).
Class III holds large heat-labile bacteriocins with a molecular mass of > 30 kDa. Their affiliation within the bacteriocins is debated, and suggestions have been made to exclude this class and rather name these proteins as bacteriolysins (Nes et al. 2007).
Class IV is comprised of circular bacteriocins which attain a closed structure after being
ribosomally synthesized as linear peptides. The linear peptide is also connected to a leader
sequence that is cleaved off before formation of the mature circular structure (figure 1.5). PTMs
different from those of the lantibiotics covalently link the two terminals of the cleaved peptide by
inducing condensation of the N- and C-terminal amino- and carboxyl-groups, resulting in a new
peptide bond. This compact structure contributes to the high stability in regard to temperature
and pH observed within this class, as well as providing resistance to a number of proteolytic
enzymes as result of reduced cleavage sites. Class IV bacteriocins generally exhibit broad
antimicrobial activity, with members presently identified ranging from 58 to 70 aa. To this date
10 circular bacteriocins have been characterized (Gabrielsen et al. 2014).
15
Figure 1.5: The structures of various class IV bacteriocins have been predicted using homology modeling with SWISS-MODEL, and are displayed here using PyMol. The positively charged residues are colored in
magenta and negatively charged residues in cyan. The point of circularization are indicated by arrows (Gabrielsen et al. 2014).
1.4 Bacteriocins addressed in this thesis.
This particular work focuses on receptor molecules for members of the leaderless class IId bacteriocins. These are produced without leader peptides at the N-terminus, and do not undergo extensive PTMs.
1.4.1 Leaderless bacteriocin LsbB
Within the leaderless bacteriocins resides a narrow-spectrum member termed LsbB, which is produced by L. lactis subsp. lactis BGMN1-5 and displays specific activity exclusively towards other L. lactis strains. A previous study has established a zinc-dependent metallopeptidase encoded by the lactococcal yvjB as receptor for LsbB (Uzelac et al. 2013).
A subsequent study (Ovchinnikov et al. 2014) elucidated questions concerning the structure-
function relationship of LsbB and its lactococcal target bacteria. CD spectroscopy was used to
determine the optimal trifluoroethanol (TFE) and dodecylphosphocholine (DCP) conditions for
obtaining maximum structuring of LsbB in solution, and NMR experiments were subsequently
carried out for structural analysis of the peptide. The results concluded through superimposing
the NMR structures with lowest energies that LsbB is divided into an α-helix region near the N-
terminus, and a more unstructured loop region by the C-terminus (figure 1.6). The amphiphilic α-
helix residues are followed by a short middle region of basic amino acids thought to facilitate the
16
initial and unspecific electrostatic interaction between the bacteriocin and phospholipid layer of the target cell membrane.
Figure 1.6: The structure of LsbB in DPC solution, represented as cartoon in Pymol from the NMR data, and showing the N-terminal region forming an alpha-helix secondary structure, with the C-terminal as a less organized loop region (Ovchinnikov et al. 2014).
Further experimental work suggested after evaluating the blocking activity of various truncated LsbB peptides that the receptor binding domain of LsbB is located in the C-terminus. Only peptides containing the C-terminus and most importantly a KxxxGxxPWE sequence motif (as from here termed the LsbB-like motif) were able to compete with LsbB by binding the receptor and inhibiting LsbB from accessing the target. In sufficiently high concentrations these truncated peptides could block the antibacterial activity completely as they do not possess antibacterial properties themselves. Additionally, alanine substitution of Trp
25(W25A) demolished all
antimicrobial activity from LsbB, proposing that this residue is particularly crucial for successful receptor binding. These results imply that the C-terminus is responsible for initiating specific binding with the YvjB receptor, followed by disruption of the membrane integrity and pore- formation ultimately killing the target bacterium.
The consensus motif of LsbB was additionally found to be conserved across several enterococcal
bacteriocins, including Ej97, EntQ and an unannotated peptide termed EntK1. Truncated C-
terminus peptides derived from these enterocins could also block the activity of LsbB, which
substantiates the involvement of this sequence motif in receptor binding, as well proposing that
these related bacteriocins also bind to the same receptor as LsbB. Curiously, these enterocins
simultaneously display a significantly broader spectrum of activity than LsbB.
17 The study also points out the apparent similarities between these broad-spectrum enterocins and narrow spectrum LsbB (all likely targeting the zinc-dependent metallopeptidase), and the broad- spectrum pediocin-like bacteriocins and narrow-spectrum lactococcin A (all targeting the man- PTS receptor). The extreme species specificity regarding lactococcal man-PTS exclusively functioning as receptor for lactococcin A has been established as a consequence of sequence differences between the lactococcal and non-lactococcal man-PTS genes. The same might be relevant for the lactococcal and non-lactococcal zinc-dependent metallopeptidases (Ovchinnikov et al. 2014).
1.4.2 Leaderless enterocins
As the LsbB-like motif (crucial for the activity) has been identified in several leaderless
enterocins, i.e. bacteriocins with enterococcal producers, these bacteriocins are also presumed to target the same receptor. But unlike LsbB, they exhibit broad activity spectra reaching far beyond lactococcal species. The work for this thesis utilized some of these bacteriocins (with emphasis on Ej97) to investigate this conundrum.
Enterocin Ej97: 44 aa peptide produced by Enterococcus faecalis strains.
Enterocin K1: 37 aa peptide produced by Enterococcus faecium strains.
Enterocin EntQ: 34 aa peptide produced by Enterococcus faecium strains.
1.4.3 Sequence similarity
Despite different activity spectra and peptide lengths, all the leaderless bacteriocins mentioned above share the conserved sequence motif KxxxGxxPWE in the C-terminal, which is essential for the activity of LsbB. W25 have proved to be particularly important in this context.
Ej97 MLAKIKAMIKKFPNPYTLAAKLTTYEINWYKQQYGRYPWERPVA 44 K1 ---MKFKFNPTGTIVKKLTQYEIAWFKNKHGYYPWEIPRC 37 EntQ ---MNFLKNGIAKWMTGAELQAYKKKYGCLPWEKISC 34 LsbB ---MKTILRFVAGYDIASHKKKTGGYPWERGKA 30
: :: :: .*:: * *** .
Figure 1.7: Multiple sequence aliment (MSA) of the sequences of Ej97, K1, EntQ and LsbB, predicted by ClustalW. Fully conserved amino acid residues are marked with “*” and less conserved residues with “:” or
“.” depending on the degree of amino acid similarity. The conserved LsbB-like motif is marked with red.
18
1.5 Bacteriocin receptors and strategies used in their identification
Due to the rapid emergence of antibiotic resistant bacteria, it is desirable to identify antimicrobial agents that target different sites than the presently utilized antibiotics in order to avoid problems with cross-resistance. The bacteriocins, targeting specific receptors in the cell membrane and exerting their activity by pore-formation are interesting in this aspect. Knowledge on their receptors for target recognition is thus desirable for further development in the field and
attracting the attention of various industries for commercial distribution. Understanding receptor- interaction is also important for engineering of improved bacteriocins and hybrids (Cotter 2014).
As will be described in this section, various strategies have been applied to elucidate receptor molecules for bacteriocins over the past decades.
The class I lantibiotic nisin and Lipid II
Nisin was the first lantibiotic to be identified, and is produced by L. lactis strains. In 1998, a precursor of peptidoglycan in the bacterial cell wall, lipid II, was identified as the docking molecule for nisin. It had already been assumed that nisin disrupts the cellular membrane of target bacteria due to exposure to nisin inducing rapid efflux of cytoplasmic components from target cells. Nisin specifically belongs to a subgroup of lantibiotics which are recognized by an elongated and cationic structure consisting of five lanthionine rings A-E. By studying the activity of mutated nisin molecules, it was discovered that the two first lanthionine rings A and B at the N-terminal are essential for binding lipid II, while the whole molecule of five rings is necessary to form lethal pores in the membrane by using the C-terminal domain. It was also observed that nisin has two mechanisms for killing target cells, either by pore formation, or by blocking lipid II from being incorporated into the peptidoglycan layer and thus inhibiting cell-wall synthesis (Hechard & Sahl 2002; Wiedemann et al. 2001). Lipid II is a cell-wall precursor of the peptidoglycan layer found in both Gram-positive and Gram-negative bacteria. Gram-positive cells acquire a thick layer of peptidoglycan, while Gram-negative cells have a thinner layer and an additional outer membrane. Nisin exhibits a broad spectrum of activity, which is reasonable as lipid II is such a common component. Several other lantibiotics such as subtilin and the
epidermin family also share this conserved ring system, suggesting that they also bind and
employ lipid II for target interaction. The epidermins do however not have the sufficient peptide
length to span bacterial membranes, and pore formation is thereby not possible. Experiments
19 have suggested that these peptides instead exert their activity by binding lipid II and inhibiting peptidoglycan synthesis by removing lipid II from its active site. This removal of lipid II from its functional site is possibly the main mechanism of lantibiotics which are too short to form stable pores (Martin & Breukink 2007).
The class IIc lactococcin A and the class IIa pediocin-like bacteriocins, and the man-PTS
Lactococcin A (class IIc) of L. lactis is exclusively active against other L. lactis species. Past
studies have revealed that lactococcin A utilizes a permeabilizing mechanism involving the
mannose phosphotransferase system (man-PTS) for target cell recognition and interaction. The
man-PTS is responsible for coupling carbohydrate import with phosphorylation of the sugar
molecules inside bacterial cells, and is not present in eukaryotes. The man-PTS consists of the
enzymes EI, HPr and the carbohydrate specific EII. EII is composed of IIAB, IIC and IID, where
the two latter are located in the membrane and involved in recognition by bacteriocins. The IIAB
component located in the cytoplasm is on the other hand not involved in sensitivity. The man-
PTS genes in different bacteria exhibit subtle sequence variations that makes it possible for the
bacteriocins to differentiate species (Kjos et al. 2009). Lactococcin A producers maintain self-
immunity by expression of a specific immunity protein LciA which form a complex with the
bacteriocin and receptor components to prevent the host from being killed. This complex appears
to occur only in the presence of the bacteriocin. By using tagged immunity-proteins (fLciA) and
the cognate bacteriocin, it was possible to co-purify a protein complex consisting of immunity
protein, receptor proteins, and bacteriocin from cell extracts by immunoprecipitation. From this
complex the man-PTS components were identified with mass spectrometry (MS) and peptide
mass fingerprinting. Additionally, lactococcin A resistance was established in strains where the
genes encoding the IIC and IID transmembrane proteins of the man-PTS had been deleted. The
study finally established that heterologous expression of the lactococcal man-PTS operon in the
naturally resistant Lactobacillus sakei rendered this species susceptible to lactococcin A. To
identify exactly which components of the man-PTS are used for target recognition by lactococcin
A, the genes were cloned and expressed as single genes or pair of genes using the NICE system
(nisin-controlled gene expression) to regulate expression. Only expression of the gene pair ptnC
(for IIC) and ptnD (for IID) conferred bacteriocin sensitivity (Diep et al. 2007). Various other
class II bacteriocins including lactococcin B (class IIc) and pediocin-like bacteriocins (class IIa)
20
also exploit the IIC and IID components of the man-PTS in killing sensitive target cells. In the case of pediocin-like class IIa bacteriocins, it has been further discovered that these bacteriocins specifically interact with a ~20 aa extracellular loop in the IIC component to bind and initiate pore-formation (Diep et al. 2007; Kjos et al. 2009; Kjos et al. 2010).
The circular class IV bacteriocin garvicin ML and the maltose ABC transporter
More recent developments in the field include the discovery of the maltose ABC transporter as receptor for the circular class IV bacteriocin garvicin ML. This bacteriocin is produced by Lactococcus garvieae DCC43, an isolate from the intestine of Mallard ducks (Borrero et al.
2011). Spontaneous L. lactis 1403 mutants with lowered sensitivity to garvicin ML were
generated and their fermentation profile was assessed for discovering possible disruption in sugar metabolism. This resulted in observations that the mutants were unable to grow effectively on starch and maltose sugar, suggesting that the mutants had lost their ability to metabolize these carbohydrates. This phenotype seemed to somehow be linked to the garvicin ML resistance and thus its receptor. Genotype analysis by sequencing revealed that the mutants had a chromosomal deletion of 13.5 kb spanning 12 open reading frames (ORFs). The data showed that this section included the genes malEFG, encoding a membrane-bound maltose ABC-transporter. Additional growth studies showed that the sensitivity to Garvicin ML was elevated when wild type cells were grown on maltose, in contrast to when the bacteria had other available resources such as glucose (Gabrielsen et al. 2012).
The two-peptide class IIb bacteriocin lactococcin G and the UppP receptor The most recent discovery in target receptor recognition was made on the two-peptide bacteriocin lactococcin G produced by certain L. lactis strains. By employing whole genome sequencing (WGS), mutations within the uppP gene were located in 12 resistant L. lactis mutants. This gene encodes a membrane spanning undecaprenyl pyrophosphate phosphatase, which functions as a component involved in peptidoglycan synthesis. Using bioinformatics tools for SNP detection, the resistant phenotype of these mutants showed correlation to mutations within or up-stream uppP. The mutants were affected by a variety of serious mutations, including premature stop codons, mutations in regulatory regions, and amino acid substitutions.
Lactococcin G is specifically potent to L. lactis strains and requires the cooperation of two
21 separate peptides (a 39 aa α-peptide ad a 35 residue β-peptide) to exert antimicrobial activity.
The importance of this gene in sensitivity was confirmed by heterologous expression of lactococcal uppP in Streptococcus pneumonia, a naturally resistant strain as lactococcin G can only target the lactococcal version of this gene. For unknown reasons, the researchers were unable to reintroduce the intact uppP gene to resistant L. lactis mutants despite several trials using various cloning vectors. Instead, the approach was modified to revolve around
heterologous expression of the lactococcal wild type uppP in Streptococcus pneumonia. This induced sensitivity to lactococcin G in the naturally resistant host. This represents the first time a receptor have been identified for a two-peptide bacteriocin. The study also revealed that another class IIb bacteriocin, enterocin 1071, also exploits UppP for target recognition. The work also included a fermentation assay to study potential differences in sugar metabolism between wild type and mutants, but did not give any positive results, implying that the receptor was not involved in sugar metabolism pathways. One mutant also had an additional mutation in oppD, the ATPase component of an ABC-type transporter system (Cotter 2014; Kjos et al. 2014).
This WGS-based receptor identification strategy (figure 1.8) proves to be a good alternative as the cost of sequencing is on the decrease, and as the heterogeneity of bacteriocins substantiates a large variety of different receptor genes remains to be identified (Cotter 2014).
Figure 1.8: Visualization of the pipeline utilized in the identification of UppP as receptor for lactococcin G.
WGS was the first step after generation of resistant mutants, and by modern SNP identification tools a candidate receptor gene was established using a bioinformatics approach. The involvement of the candidate gene was subsequently confirmed by experimental evidence through heterologous expression of the wild type gene, which introduced sensitivity to lactococcin G in a naturally resistant host (Cotter 2014).
The leaderless class IId bacteriocin LsbB and a Zn-dependent metallopeptidase
Another recent finding in bacteriocin receptors came with the identification of a Zn-dependent
metallopeptidase coded by the yvjB-gene in L. lactis as receptor for the leaderless class IId
22
bacteriocin LsbB. YvjB is a lactococcal member of the RIP metallopeptidase RseP protein family and the M50 superfamily. The strategy employed was to create a cosmid library of an LsbB-sensitive strain and experiment with which fragment from this library restored sensitivity to the bacteriocin when expressed in resistant mutants. Each cosmid library contained a fragment and the lambda phage cos-sequence to make it suitable as a cloning vector. The results showed that a cosmid carrying a 40 kb insert was able to restore sensitivity, and further subcloning of this cosmid to narrow the candidate gene possibilities revealed an open reading frame on a 1.9 kb fragment as a potential receptor gene. This open reading frame encodes the zinc-dependent protease enzyme YvjB. The involvement of this particular gene as receptor was confirmed by additional whole genome sequencing showing yvjB mutations in all mutants, deletion of functional yvjB inducing resistant phenotype, and homologous expression of wild type yvjB in naturally resistant hosts introducing sensitivity to LsbB (Uzelac et al, 2013).
Figure 1.9: A depiction of the presently identified bacteriocin receptors initiating pore-formation in
susceptible target cells. Class IIa and some IIc bacteriocins exploits components of the man-PTS, nisin and
other lantibiotics target lipid II, circular class IV garvicin ML attacks the Maltose ABC transporter, class IId
leaderless bacteriocin LsbB bind a zinc-dependent metallopeptidase, and finally recent developments have
established UppP as docking molecule for the two-peptide class IIb bacteriocin lactococcin G (Cotter 2014).
23
1.6 The zinc-dependent metallopeptidase Rsep of the M50-superfamily
As previously elaborated, lactococcal RseP (a zinc-dependent metallopeptidase YvjB) of the M50 superfamily has been identified as the receptor for the leaderless bacteriocin LsbB.
Members of this family exert proteolytic activity, and such protease enzymes serve either as cellular housekeepers by removing damaged proteins in the cell, or as part of a regulatory pathway by cleaving specific regulatory components. The latter are additionally termed RIP (regulated intramembrane proteolysis) enzymes. RseP is such a RIP enzyme.
In studies on E. coli, the rseP gene is well established as involved in initiating stress responses by cleaving an anti-sigma factor and thereby allowing cognate sigma factors to induce
transcription of stress response genes (figure 1.10). RseP is in this manner implemented in regulatory proteolysis by a proteolytic cascade reaction which is initiated in response to
extracellular stress signals. By default, the sigma factor σ
Eis bound to its regulatory anti-sigma factor RseA. Stress signals lead to the RseP cleaving the anti-sigma factor RseA in accordance with cleavage by the inner membrane protease DegS. These cleavages release the cytoplasmic domain of RseA into the cell where another protease ClpXP degrades the remaining parts of the RseA, and thus frees σ
E. This enables σ
Eto bind DNA and RNA polymerase in the promotor site upstream the genes responsible for stress response. The process is regulated in this manner to inhibit unnecessary production of stress response genes when there is no stress present cellular environment. The rseP has orthologous genes in various species (Akiyama et al. 2004; Kroos &
Akiyama 2013).
Figure 1.10: DegS cleaves the periplasmic C-terminal domain of anti-sigma factor RseA, before RseP cleaves
the trans membrane region (displayed by arrows), which release the N-terminal domain of RseA and the
connected σ
Einto the cell, where the remaining RseA is finally degraded by ClpXP in the cytoplasm. Thus,
releasing σ
Eto initiate transcription of stress response genes (Kroos & Akiyama 2013).
24
1.7 Aim of this study
The main goal of this thesis was to identify the target receptor for the leaderless bacteriocin Ej97.
The intriguing fact that sequence characteristics of Ej97 substantiates that it likely exploits the LsbB-receptor, while simultaneously possessing the ability to dock with many other bacterial species LsbB itself is unable to attack, made this an interesting task.
The work completed in this study centered on the following tasks:
1. A spot-on-lawn inhibition spectrum assay for evaluating the activity of three homologous bacteriocins LsbB, K1 and Ej97 towards approximately 50 indicator bacteria. Indicators with naturally resistant colonies occurring within the inhibition zones would be the basis of further work progress.
2. Accumulation of resistant mutants and utilization of microtiter plate assays with two-fold bacteriocin dilutions to assess the resistance level (phenotype) of these mutants relative to the wild type.
3. Sequencing of the putative receptor RseP, recognized as a membrane-associated
metalloprotease and homologous with LsbB receptor YvjB in L. lactis IL1403, in order to identify mutations (genotype) causing the observed resistant phenotypes.
4. Whole genome sequencing for uncovering reasons for the observed resistance in isolates lacking mutations within rseP, and assessment of possible unspecific resistance
mechanisms present in these resistant mutants.
25
2. MATERIALS
2.1 Growth media and agars
Medium Supplier
BHI (Brain-Heart-Infusion) Oxoid
Growth medium: 9.25 g BHI dH
2O to 250 ml
Sterilized in autoclave for 15 min at 121 °C Soft agar: BHI medium with agar (8 g/L) Agar: BHI medium with agar (15 g/L)
MRS (de Man, Rogosa, Sharpe) Oxoid
Growth medium: 13 g MRS dH
2O to 250 ml
Sterilized in autoclave for 15 min at 121 °C Soft agar: MRS medium with agar (8 g/L) Agar: MRS medium with agar (15 g/L)
2.2 Bacterial strains
Table 2.1: Bacterial strains implemented in the work for this thesis.
Bacterium LMG lab strain reference code
Bacillus cereus
LMG 2805
Enterococcus avium 208
LMG 3465
Enterococcus faecalis UI50
LMG 2333
E. faecalis 2708 RA
LMGT 3358*
E. faecalis DEC23
LMGT 3386
E. faecalis SMF37
